Evidence from Accreted Seamounts for a Depleted Component in the Early Galapagos Plume

Evidence from accreted seamounts for a depleted component in the early Galapagos plume

David M. Buchs1, Kaj Hoernle2,3, Folkmar Hauff2, and Peter O. Baumgartner4 1School of Earth and Ocean Sciences, Cardiff University, Main Building, Cardiff CF10 3AT, UK 2GEOMAR Helmholtz Centre for Ocean Research Kiel, Wischhofstrasse 1-3, 24148 Kiel, Germany 3Christian-Albrechts University of Kiel, 24118 Kiel, Germany 4Institute of Earth Sciences, University of Lausanne, Bâtiment Géopolis, 1015 Lausanne, Switzerland

ABSTRACT that ca. 90 Ma depleted komatiites from Gor- The existence of an intrinsic depleted component in mantle plumes has previously been gona Island (Kerr et al., 1995) and ca. 80 Ma proposed for several hotspots in the Pacific, Atlantic, and Indian Oceans. However, formation depleted basalts from Ocean Drilling Program of these depleted basalts is often associated with unusual tectonomagmatic processes such (ODP) Site 1001 in the Caribbean large igneous as plume-ridge interaction or multistage melting at plume initiation, where depleted basalts province (CLIP) (Kerr et al., 2009) provide addi- could reflect entrainment and melting of depleted upper mantle. Late Cretaceous to middle tional evidence for the existence of an intrinsic Eocene seamounts that accreted in Costa Rica and are part of the early Galapagos hotspot depleted component in the Galapagos hotspot. track provide new insights into the occurrence and nature of intrinsic depleted components. However, Site 1001 basalts formed on the top of The Paleocene (ca. 62 Ma) seamounts include unusually depleted basalts that erupted on the the CLIP by second-stage melting of a mantle Farallon plate far from a mid-ocean ridge. These basalts closely resemble Gorgona komatiites plume source that had already been melted to in terms of trace element and radiogenic isotope composition, suggesting formation from a form the base of the CLIP (Kerr et al., 2009), similar, refractory mantle source. We suggest that this source may be common to plumes, thus making a direct link with plume magmatism but is only rarely sampled due to excessive extents of melting required to extract melts from questionable. Gorgona komatiites are arguably the most refractory parts of a heterogeneous mantle plume. the best evidence for the existence of an intrinsic depleted plume component; however, these rocks INTRODUCTION also recycled, thus forming an intrinsic depleted are very unusual, without known equivalents in It is generally assumed that ocean-island component in mantle plumes (Kerr et al., 1995). the Mesozoic, and their exact provenance and basalts (OIBs) are derived from mantle plumes Basalts with depleted incompatible element link to the nearby Galapagos hotspot is unclear and that these melts commonly have more and isotopic compositions have been found at (Kerr and Tarney, 2005). enriched compositions than mid-oceanic ridge numerous hotspots, e.g., Iceland (Fitton et al., Due to ambiguities concerning the origins basalts (MORBs). The most widely accepted 1997; Thirlwall et al., 2004), Galapagos (White of depleted basalts formed in oceanic plateaus explanation for the enriched composition of et al., 1993), and Hawaii (Keller et al., 2000). or at hotspots where a plume interacts with a hotspot lavas is that plumes carry recycled mate- The depleted basalts are most common when mid-ocean ridge, it remains essential to provide rial such as altered oceanic crust and recycled the plume is located near a mid-ocean ridge. additional constraints on the possible existence sediment, which can be stored for tens of mil- The origin of the depleted component, however, of an intrinsic depleted component in mantle lions to billions of years in the mantle before remains controversial in most cases due to the plumes. Novel support for the existence of this returning to the surface, where the recycled similarity in composition between melts from component is provided here by new geochemi- material melts preferentially relative to depleted entrained depleted upper mantle and from a cal data from Late Cretaceous to middle Eocene upper mantle to form OIBs (Hofmann and depleted component intrinsic in the plume. Both accreted seamounts in the Osa Igneous Complex White, 1982). If the altered upper crust and sedi- will melt due to shallow upwelling and high- (Costa Rica), which formed at the early Galapa- ment of the oceanic lithosphere are recycled into degree melting of the plume beneath a ridge gos hotspot far from a mid-ocean ridge. the mantle, then it is likely that the lower, unal- potentially coupled with previous extraction of tered depleted crust and lithospheric mantle are enriched melts at depth. It has been proposed GEOLOGICAL BACKGROUND AND METHODS CENTRAL 100°W 90°W AMERICA 80°W The Osa Igneous Complex (OIC) is exposed CARIBBEAN PLATE on the Osa and Burica Peninsulas at the south- 10°N COCOS 10°N PLATE western edge of the CLIP (Fig. 1; Fig. DR1 East Pacific 1 Rise (EPR) in the GSA Data Repository ). The complex

E Figure 1. Tectonic map of e includes an assemblage of Cretaceous to Eocene

AT the central-eastern Pacific Coiba Ridge Ocean. Black areas in oceanic sequences predominantly composed of crust formed at the EPR Cocos Ridg Central America represent accreted oceanic crust formed at the CNSC 1 CIFIC PL Malpeloe GSA Data Repository item 2016125, data tables, Cocos-Nazca Spreading Ridg sequences; square out- PA Figures DR1–DR3 (geological map and geochemical Center (CNSC) lines the location of Figure 0° 0° DR1 (see footnote 1). figures), and additional comments on tectonostratig- SOUTH raphy, analytical methods, data selection, and isotope NAZCA AMERICAN Galapagos data, is available online at www.geosociety.org/pubs PLATE Carnegie PLATE Islands / Hotspot Ridge /ft2016.htm, or on request from editing@geosociety .org or Documents Secretary, GSA, P.O. Box 9140, 100°W 90°W 80°W Boulder, CO 80301, USA.

Downloaded from http://pubs.geoscienceworld.org/gsa/geology/article-pdf/44/5/383/3549647/383.pdf by guest on 30 September 2021 3 Accreted basalts Figure 2. Selected geochemi- A B y Cocos and Group 1 (enriched) Carnegie Ridges Group 2 (intermediate) cal characteristics of igneous 1 accreted Group 3 (depleted) accreted OIBs OIBs rocks from the outer Osa Igne-

N south Central Genovesa EPR off-axis EPR off-axis ous Complex (small diamonds American Icelandic plume arra Island seamounts Cocos and oceanic plateau seamounts Carnegie Ridges are samples from the inner

Nb/Y Osa Igneous Complex). Data

[Dy/Yb ] Gorgona komatiites sets used for comparison 1 0.1 Boninites Boninites include whole-rock and glass Site MORB MORB 1001 analyses from representative

south Central America settings. EPR—East Pacific oceanic plateau Genovesa Gorgona Site Island Rise; MORB—mid-oceanic ridge komatiites 1001 0.5 0.01 basalt; accreted OIBs—accreted 0.21 6 11Zr/Y 0 [La/Sm]N ocean-island basalts in Central 10 America (for references, see C D the Data Repository [see foot- Cocos and Tholeiitic Alkaline y Carnegie Ridges OIB array note 1]; all selected samples y Genovesa Island accreted have MgO > 5.5 wt%). Site 1001 arra EPR off-axis 1 south Central American 1 seamounts OIBs volcanic arc arra oceanic plateau b MORB is from the Ocean Drilling Pro- b MORB-OIB gram. A: (Zr/Y) versus (Nb/Y)

2 N accreted BoniniteBoninites OIBs MORB array diagram (N—primitive mantle Th/Y Cocos and N TiO /Y 0.1 Genovesa EPR off-axis Carnegie Ridges normalized). B: Zr/Y versus Nb/Y Gorgona Island komatiites seamounts diagram (after Fitton et al., 1997). south Central American Gorgona Site oceanic plateau C: Nb/Yb versus Th/Yb diagram Site komatiites 1001 1001 (after Pearce, 2008). D: Nb/Yb MORB 0.01 0.1 versus TiO /Yb diagram (after 0.11 10 100 2 Nb/Yb 0.11Nb/Yb 10 100 Pearce, 2008).

pillow and massive basalt flows with subordi- additionally indicated by loss on ignition values pelagic sedimentary rocks in igneous sequences nate gabbro (locally pegmatitic) and thin (<10 of 0.55–5.45 wt% (Table DR2). As a conse- of this group unambiguously support formation m thick) interbeds of pelagic and volcanicla- quence, the origin of the outer OIC is determined in a mid-ocean ridge or intraplate setting. The stic sedimentary rocks (e.g., Di Marco et al., based on a combination of immobile trace ele- trace element composition of group 1 closely 1995; Hauff et al., 2000; Hoernle et al., 2002; ment discrimination diagrams (Fitton et al., resembles that of seamount chains formed at Buchs et al., 2009). The complex has been sub- 1997; Pearce, 2008), radiogenic isotope ratios the Galapagos hotspot since the Miocene (Fig. divided into an inner OIC interpreted as a Late that are relatively insensitive to alteration, geo- 2). Isotope compositions are clearly enriched Cretaceous (ca. 85 Ma) oceanic plateau and an chemical comparison with a selection of possible compared to those of East Pacific Rise (EPR) outer OIC including several accreted seamounts volcanic analogues, field observations, and exist- MORBs (Fig. 3). This suggests that group 1 ranging in age between the Late Cretaceous (ca. ing regional constraints. New geochemical data could represent late-stage volcanic sequences 85 Ma) and middle Eocene (ca. 41 Ma) (Buchs from the inner OIC confirm an oceanic plateau et al., 2009). Tectonostratigraphic constraints origin; they have a composition indistinguish-

define pre–late Eocene accretion ages for the able from contemporaneous Late Cretaceous 39.0 outer OIC (see the Data Repository). Here we oceanic plateau sequences observed elsewhere A Central America 38.5 oceanic plateau report new geochemical data that confirm an in south Central America (Hauff et al., 2000; Gorgona 38.0 komatiites

oceanic plateau origin for the inner OIC; our Hoernle et al., 2002; Buchs et al., 2010). These Pb Genovesa main focus is on the poorly constrained origin data are reported here only for comparison with 20 4 37.5 island average ingrowth correction (Group 3) Pb / of the outer OIC. the outer OIC (Fig. 2; Figs. DR2 and DR3). 37.0 20 8 Site Accreted basalts 1001 In order to determine the composition of In contrast to oceanic plateau sequences of Group 1 (enriched) 36.5 EPR MORB the OIC, 54 whole-rock samples of basalt and the inner OIC, accreted igneous rocks of the Group 2 (intermediate) noncumulative gabbro were selected for X-ray outer OIC display large compositional heteroge- Group 3 (depleted) Site fluorescence and laser ablation–inductively cou- neity, with three geochemical groups defined by 12 B EPR MORB 1001 Gorgona pled plasma–mass spectrometry analyses at the distinct immobile trace element contents (Fig. 11 komatiites University of Lausanne (Switzerland). Nd-Pb DR3). Group 1 has an enriched (OIB like) signa- 10 average ingrowth correction (Group 3)

isotope data on a subset of 7 samples from the ture with (La/Sm)N = 0.91–1.94 (N is primitive d Genovesa 9 island outer OIC were analyzed at the GEOMAR mantle normalized), (Dy/Yb) = 1.19–1.53, and εN N 8 Helmholtz Center (Kiel, Germany). Results, full transitional tholeiitic to alkaline affinities (Fig. Central America 7 oceanic plateau analytical methods, and detailed evaluation of 2; Figs. DR2 and DR3). Two samples have εNd 206 204 6 whether initial or measured radiogenic isotopes = 6.7–8.3 and Pb/ Pb = 19.2–19.4 (Fig. 3). 17.0 17.5 18.0 18.5 19.0 19.5 20.0 provide the least biased source information are Basalts of this group occur predominantly close 206Pb/204Pb provided in the Data Repository, in and Tables to boundaries of the outer OIC and are locally 206 204 DR1–DR4 therein. in contact with the Osa Mélange and inner Figure 3. Thorogenic and Pb/ Pb versus eNd isotope correlations of Osa basalts. Aver- OIC (Fig. DR1). Pelagic limestones interbed- age 62 m.y. radiogenic ingrowth vector for RESULTS ded with pillow lavas and peperites of group group 3 basalts is minor compared to overall Analyzed basalts and gabbros have only been 1 define Campanian to middle Eocene ages of variability of the data and reference fields. exposed to low-grade metamorphism. Low to formation (Buchs et al., 2009). The absence of EPR—East Pacific Rise; MORB—mid-oceanic ridge basalt. For data sources of reference moderate hydrothermal alteration, however, is suprasubduction geochemical characteristics fields and a discussion on the effects of age frequent with local replacement of glass, olivine, (e.g., low Nb and Th/Yb for a given Nb/Yb correction, see the Data Repository (see foot- and feldspar by secondary phases. Alteration is ratio; Fig. 2) and the occurrence of interbedded note 1).

384 www.gsapubs.org | Volume 44 | Number 5 | GEOLOGY

Downloaded from http://pubs.geoscienceworld.org/gsa/geology/article-pdf/44/5/383/3549647/383.pdf by guest on 30 September 2021 of seamounts, which accreted before and after (Werner et al., 2003; Harpp et al., 2005; Fig. 2). Mesozoic MORB compositions, which have emplacement of the bulk of the outer OIC. Unlike MORB and small off-axis seamounts that notably less radiogenic Nd at a given 206Pb/204Pb Most of the outer OIC includes an assem- only form minor topographic anomalies, larger (Fig. 3B), thus ruling out significant involve- blage of group 2 and 3 igneous rocks interpreted volcanic structures in a hotspot track are more ment of a depleted upper mantle source in the from field relationships as accreted seamounts likely to be accreted. In addition, the Galapagos formation of group 3 basalts. Instead, the incom- with moderately to very depleted compositions. hotspot has been the sole source of intraplate patible element and isotopic composition of A seamount origin is also in good agreement volcanism in the central-eastern Pacific since these basalts is very similar to that of Gorgona with geochemical constraints (see following) the Late Cretaceous (e.g., Pindell and Kennan, komatiites. We conclude therefore that melting and the paucity of sedimentary rocks found in 2009). Numerous studies along the Central of an intrinsic depleted plume component is association with group 2 and group 3 sequences. American forearc support continuous formation required to account for the formation of depleted This latter point is a common characteristic of of oceanic islands and large seamounts in the Paleocene seamounts found in Costa Rica. This accreted seamounts that clearly contrasts with central-eastern Pacific since the Late Cretaceous is a significant result that not only reveals the accreted MORB sequences that are generally and their subsequent accretion to the Central existence of an intrinsic depleted component in associated with thick pelagic-hemipelagic sedi- American margin (Hauff et al., 2000; Hoernle et the early Galapagos plume, but also shows that mentary deposits (e.g., Kusky et al., 2013). al., 2002; Buchs et al., 2011; Trela et al., 2015). this component contributed to formation of the Group 2 igneous rocks have tholeiitic affin- The preceding observations provide strong sup- earliest Galapagos hotspot tracks. ities with intermediate incompatible element port for the formation of the outer OIC from the compositions, characterized by nearly flat mul- older Galapagos hotspot track (ca. 85–41 Ma) Implications for Plume Magmatism tielement patterns, except for slightly lower (Fig. 4). The outer OIC therefore serves as an The existence of an intrinsic depleted compo- Nb and Th primitive mantle–normalized con- important archive of the nature and dynamics nent in mantle plumes was previously proposed

tents with (La/Sm)N = 0.76–0.98, (Dy/Yb)N = of the early Galapagos plume, shortly after final based on depleted basalts in oceanic islands and/

0.97–1.12, and (Th/Yb)N = 0.37–0.67 (Fig. 2; emplacement ca. 90 Ma of the CLIP above the or seamounts of the Iceland, Galapagos, and Figs. DR2 and DR3). The trace element com- starting plume head of the Galapagos hotspot. Hawaii hotspots (e.g., Kerr et al., 1995; Hoernle position of these rocks overlaps with that of et al., 2000; Keller et al., 2000) and depleted enriched MORB, oceanic plateau sequences Depleted Component in the Early komatiites in oceanic plateau sequences of of the inner OIC, and the Cocos and Carnegie Galapagos Plume Gorgona Island (Kerr et al., 1995). In all these Ridges (Fig. 2). A sample from group 2 has Paleocene (group 3) basalts accreted in the depleted ε (8.9) and 206Pb/204Pb (18.5) and outer OIC exhibit one of the most depleted Nd Location of Central plots close to the compositional field of EPR incompatible element compositions and Nd WE American Arc at ≤75 Ma Antilles Arc MORB and Genovesa Island, Galapagos Archi- isotopic compositions found to date in oce- pelago (Fig. 3). Although a MORB origin cannot anic basalts, in particular in those from hotspot CLIP be totally excluded based on geochemical data tracks, suggesting unusual melting conditions alone, a seamount origin is in better agreement and/or petrogenetic processes in the early Gala- with lithostratigraphic observations. Radiolarite pagos plume. Two main petrogenetic models, Initiating Galapagos interbedded with the lavas constrain the age of exemplified by intraplate oceanic igneous rocks plume tail will remain formation of group 2 to the Coniacian–Santo- from the Caribbean and central-eastern Pacific A stationary nian (ca. 85 Ma; Buchs et al., 2009). (Fig. 2), could account for formation of depleted ca. 90 Ma Group 3 igneous rocks have tholeiitic affinities basalts at hotspots: (1) plume-ridge interaction WEpaleo-hotspot Central ODP (Fig. 2D; Fig. DR2) with an extreme depletion in associated with entrainment of MORB astheno- track American Arc Site 1001 the most incompatible trace elements [(La/Sm)N sphere in a hot mantle plume (White et al., 1993; = 0.27–0.61; (Dy/Yb) = 0.95–1.08; (Th/Yb) = Harpp et al., 2002) or melting of an intrinsic CLIP N N Farallon 0.09–0.23] (Fig. 2A). Incompatible trace element plume component (Hoernle et al., 2000), e.g., Caribbean contents of this group resemble those of depleted Genovesa Island and Genovesa ridge in the Gala- early komatiites from Gorgona Island. Isotope data for pagos Archipelago; and (2) high-temperature or Galapagos hotspot Formation five samples (including sample OS6 from Hauff hydrous melting of a depleted, intrinsic mantle of Group 3 et al. 2000) have depleted εNd (10.6–11.4) and plume component during the earliest stages of B basalts ca. 62 Ma 206Pb/204Pb (18.5–18.7) distinct from EPR MORB plume volcanism, e.g., Gorgona komatiites (Kerr but similar to Gorgona komatiites (Fig. 3), et al., 1995; Kamenetsky et al., 2010). WECocos Central American Arc indicating the existence of unusually depleted An origin through plume-ridge interaction CNSC Ridge oceanic basalts in the outer OIC (group 3). The involving melting of the upper mantle due to CLIP age of formation of group 3 is Paleocene, based increased heat from the plume and very shal- Nazca Cocos on a 62.1 ± 0.8 Ma age determined by 40Ar/39Ar low upwelling is not likely because break-up dating (Hauff et al., 2000). of the Farallon plate to form the Cocos-Nazca plume-ridge Galapagos magmatism spreading center did not occur until 23 Ma hotspot DISCUSSION (Barckhausen et al., 2008), i.e., nearly 40 m.y. C after formation of the accreted seamounts (Fig. ca. 23 Ma Support for a Galapagos Provenance 4). There is no evidence for the presence of a Figure 4. The evolution of the Galapagos A paleo–Galapagos hotspot origin for the spreading center in the vicinity of the Galapa- plume. A: Initiation. B: Migration of the Carib- Cretaceous to Paleocene seamounts accreted gos hotspot before 23 Ma (Pindell and Kennan, bean large igneous province (CLIP) away from in the OIC is in good agreement with regional 2009). In addition, incompatible element ratios the early Galapagos hotspot. ODP—Ocean Drilling Program C: Present-day situation, fol- tectonic, age, and lithostratigraphic constraints. for group 3 basalts largely are outside of the lowing break-up of the Farallon plate into the The compositional range of the three groups MORB field (Fig. 2) and radiogenic isotope Cocos and Nazca plates 23 Ma. CNSC—Cocos resembles that of the Galapagos hotspot tracks compositions do not overlap with modern and and Nazca spreading center.

GEOLOGY | Volume 44 | Number 5 | www.gsapubs.org 385

Downloaded from http://pubs.geoscienceworld.org/gsa/geology/article-pdf/44/5/383/3549647/383.pdf by guest on 30 September 2021 settings, a depleted component has been asso- the Caribbean plate: Geological Society of Lon- Keller, R.A., Fisk, M.R., and White, W.M., 2000, Iso- ciated with particular melting conditions due don Special Publication 328, p. 411–456, doi:10 topic evidence for Late Cretaceous plume-ridge .1144/SP328.17. interaction at the Hawaiian hotspot: Nature, to plume-ridge interaction (Iceland, Galapagos, Buchs, D.M., Arculus, R.J., Baumgartner, P.O., v. 405, p. 673–676, doi:10 .1038/35015057. Hawaii), formation of an oceanic plateau above Baumgartner-Mora, C., and Ulianov, A., 2010, Kerr, A.C., and Tarney, J., 2005, Tectonic evolution of a starting plume head (Gorgona Island), or Late Cretaceous arc development on the SW the Caribbean and northwestern South America: remelting of plume material after previous melt margin of the Caribbean plate: Insights from the The case for accretion of two Late Cretaceous Golfito, Costa Rica, and Azuero, Panama, com- oceanic plateaus: Geology, v. 33, p. 269–272, doi: extraction (ODP Site 1001 in the Caribbean). plexes: Geochemistry, Geophysics, Geosystems, 10.1130/G21109.1. Because they did not form over a plume head or v. 11, Q07S24, doi:10.1029/2009GC002901. Kerr, A.C., Saunders, A.D., Tarney, J., Berry, N.H., through plume-ridge interaction, group 3 basalts Buchs, D.M., Arculus, R.J., Baumgartner, P.O., and and Hards, V.L., 1995, Depleted mantle-plume from the outer OIC represent a new occurrence Ulianov, A., 2011, Oceanic intraplate volcanoes geochemical signatures: No paradox for plume exposed: Example from seamounts accreted in theories: Geology, v. 23, p. 843–846, doi:10 of plume-related depleted basalts that formed Panama: Geology, v. 39, p. 335–338, doi:10.1130 .1130/0091-7613(1995)023<0843:DMPGSN>2 ~30 m.y. after the peak of plume head volca- /G31703.1. .3.CO;2. nism that formed the Caribbean large igneous Di Marco, G., Baumgartner, P.O., and Channell, Kerr, A.C., Pearson, D.G., and Nowell, G.M., 2009, province at ca. 90 Ma (Fig. 4). J.E.T., 1995, Late Cretaceous–early Tertiary Magma source evolution beneath the Caribbean oceanic plateau: new insights from elemental and As shown by Herzberg and Gazel (2009), paleomagnetic data and a revised tectonostratigraphic subdivision of Costa Rica and western Sr-Nd-Pb-Hf isotopic studies of ODP Leg 165 early Galapagos lavas with ages of ca. 65 Ma Panama; geologic and tectonic development of Site 1001 basalts, in James, K.H., et al., eds., The were formed at potential temperatures of ≥1500 the Caribbean plate boundary in southern Central origin and evolution of the Caribbean plate: Geo- °C, whereas lavas younger than 15 Ma belonging America, in Mann, P., ed., Geologic and tectonic logical Society of London Special Publication to the Galapagos Archipelago and hotspot tracks development of the Caribbean plate boundary in 328, p. 809–827, doi:10.1144/SP328.31. southern Central America: Geological Society Kusky, T.M., Windley, B.F., Safonova, I., Wakita, K., formed at lower potential temperatures (1400– of America Special Paper 295, p. 1–28, doi:10 Wakabayashi, J., Polat, A., and Santosh, M., 2013, 1500 °C). Although melting of pyroxenite could .1130/SPE295-p1. Recognition of ocean plate stratigraphy in accre- occur at lower melting temperatures (Trela et al., Fitton, J.G., Saunders, A.D., Norry, M.J., Hardarson, tionary orogens through Earth history: A record of 2015), we speculate that hotter conditions in the B.S., and Taylor, R.N., 1997, Thermal and chem- 3.8 billion years of sea floor spreading, subduc- ical structure of the Iceland plume: Earth and tion, and accretion: Gondwana Research, v. 24, early hotspot setting promoted melting of the Planetary Science Letters, v. 153, p. 197–208, p. 501–547, doi:10.1016/j.gr.2013.01.004. most refractory component of a heterogeneous doi:10.1016/S0012-821X(97)00170-2. Pearce, J.A., 2008, Geochemical fingerprinting of Galapagos plume to form group 3 basalts. If this Harpp, K.S., Wirth, K.R., and Korich, D.J., 2002, oceanic basalts with applications to ophiolite is correct, this component could be ubiquitous Northern Galápagos Province: Hotspot-induced, classification and the search for Archean oceanic near-ridge volcanism at Genovesa Island: Geol- crust: Lithos, v. 100, p. 14–48, doi:10 .1016 /j .lithos in mantle plumes, but is only rarely sampled in ogy, v. 30, p. 399–402, doi:10 .1130 /0091 -7613 .2007.06 .016. significant proportions in ocean basalts due to (2002)030<0399:NGPPHI>2.0.CO;2. Pindell, J., and Kennan, L., 2009, Tectonic evolution particular petrogenetic conditions required to Harpp, K.S., Wanless, V.D., Otto, R.H., Hoernle, of the Gulf of Mexico, Caribbean and northern melt refractory sources. The exact nature and K.A.J., and Werner, R., 2005, The Cocos and South America in the mantle reference frame: Carnegie aseismic ridges: A trace element record An update, in James, K., et al., eds., The origin compositional variability of this component of long-term plume-spreading center interaction: and evolution of the Caribbean plate: Geologi- remains to be more fully investigated through Journal of Petrology, v. 46, p. 109–133, doi:10 cal Society of London Special Publication 328, a systematic comparison of depleted basalts at .1093/petrology/egh064. p. 1–55, doi:10.1144/SP328.1. hotspots and mid-ocean ridges globally. How- Hauff, F., Hoernle, K.A., van den Bogaard, P., Alva Thirlwall, M.F., Gee, M.A.M., Taylor, R.N., and Mur- ton, B.J., 2004, Mantle components in Iceland and ever, due to relatively limited occurrence of rado, G.E., and Garbe-Schonberg, D., 2000, Age and geochemistry of basaltic complexes in west- adjacent ridges investigated using double-spike depleted basalts at modern hotspots, the study ern Costa Rica: Contributions to the geotectonic Pb isotope ratios: Geochimica et Cosmochimica of ancient OIBs will remain essential in helping evolution of Central America: Geochemistry, Acta, v. 68, p. 361–386, doi:10.1016 /S0016 -7037 characterize this component. Geophysics, Geosystems, v. 1, 1009, doi:10 .1029 (03)00424-1. /1999GC000020. Trela, J., Vidito, C., Gazel, E., Herzberg, C., Class, C., Herzberg, C., and Gazel, E., 2009, Petrological Whalen, W., Jicha, B., Bizimis, M., and Alvarado, ACKNOWLEDGMENTS evidence for secular cooling in mantle plumes: Na- G.E., 2015, Recycled crust in the Galápagos We thank Andrew Kerr, reviewers David Peate and ture, v. 458, p. 619–622, doi:10.1038 /nature07857. Plume source at 70 Ma: Implications for plume William M. White, an anonymous reviewer, and editor Hoernle, K.A., Werner, R., Phipps Morgan, J., Bryce, evolution: Earth and Planetary Science Letters, Brendan Murphy for their insightful and constructive J., and Mrazek, J., 2000, Existence of a complex v. 425, p. 268–277, doi:10.1016/j.epsl.2015.05 comments. Alexey Ulyanov and Goran Andji at the ć spatial zonation in the Galapagos plume: Geol- .036. University of Lausanne and Silke Hauff at GEOMAR ogy, v. 28, p. 435–438, doi:10 .1130 /0091 -7613 Werner, R., Hoernle, K., Barckhausen, U., and Hauff, provided logistical and analytical support. This study (2000)28<435:EOCSZI>2.0.CO;2. F., 2003, Geodynamic evolution of the Galápagos was supported by grants 00021–105845 and 200021– Hoernle, K., van den Bogaard, P., Werner, R., Lissinna, hot spot system (central east Pacific) over the past 105845 of the Swiss National Science Foundation and B., Hauff, F., Alvarado, G., and Garbe-Schon- 20 m.y.: Constraints from morphology, geochem- GEOMAR funds to cover isotope analyses. berg, D., 2002, Missing history (16–71 Ma) of the istry, and magnetic anomalies: Geochemistry, Galapagos hotspot: Implications for the tectonic Geophysics, Geosystems, v. 4, 1108, doi:10 .1029 REFERENCES CITED and biological evolution of the Americas: Geol- /2003GC000576. Barckhausen, U., Ranero, C.R., Cande, S.C., Engels, ogy, v. 30, p. 795–798, doi:10 .1130 /0091 -7613 White, W.M., McBirney, A.R., and Duncan, R.A., M., and Weinrebe, W., 2008, Birth of an intraoce- (2002)030<0795:MHMOTG>2.0.CO;2. 1993, Petrology and geochemistry of the Galá- anic spreading center: Geology, v. 36, p. 767–770, Hofmann, A.W., and White, W.M., 1982, Mantle pagos Islands: Portrait of a pathological mantle doi:10.1130/G25056A.1. plumes from ancient oceanic crust: Earth and plume: Journal of Geophysical Research, v. 98, Buchs, D.M., Baumgartner, P.O., Baumgartner-Mora, Planetary Science Letters, v. 57, p. 421–436, doi: p. 19,533–19,563, doi:10.1029/93JB02018. C., Bandini, A.N., Jackett, S.-J., Diserens, M.-O., 10.1016/0012-821X(82)90161-3. and Stucki, J., 2009, Late Cretaceous to Miocene Kamenetsky, V.S., Gurenko, A.A., and Kerr, A.C., 2010, Manuscript received 18 December 2015 seamount accretion and mélange formation in Composition and temperature of komatiite melts Revised manuscript received 28 March 2016 the Osa and Burica Peninsulas (southern Costa from Gorgona Island, Colombia, constrained from Manuscript accepted 28 March 2016 Rica): Episodic growth of a convergent margin, in olivine-hosted melt inclusions: Geology, v. 38, James, K., et al., eds., The origin and evolution of p. 1003–1006, doi:10 .1130 /G31143 .1. Printed in USA

386 www.gsapubs.org | Volume 44 | Number 5 | GEOLOGY

Downloaded from http://pubs.geoscienceworld.org/gsa/geology/article-pdf/44/5/383/3549647/383.pdf by guest on 30 September 2021